Attractiveness Test
An attractiveness test that hands you one number and no explanation is closer to a slot machine than a measurement. This one is built the other way round. It measures the facial proportions that appear most often in attractiveness research, shows you every number it took, and is explicit about which ones carry weight and which are classical convention.
About the Attractiveness Test
The measurements come from a 478-point 3D face mesh mapped onto your photo by an on-device AI model. Symmetry carries the heaviest weight at 25%, compared across nine landmark pairs, because left-right symmetry has the strongest research support of anything measurable from a photo. Facial thirds and your length-to-width ratio against the classical 1.618 reference add another 25%. Eye spacing and canthal tilt contribute 20%, jawline taper and chin angle 15%, and cheekbone prominence 15%. Every weight and every raw ratio is printed on your results page, alongside a pair-by-pair symmetry breakdown and mirrored composites of each side of your face.
Here is the part most attractiveness tests leave out: geometry is a minority shareholder in how attractive anyone actually looks. Expression, grooming, skin, styling, context, and familiarity all matter enormously, and none of them can be measured from a single still photo. Studies testing proportion formulas against rated attractiveness return mixed results. Treat your score as a description of your proportions and a starting point for styling decisions, not a ranking of you. Everything runs inside your browser, your photo is never uploaded, and the tool is for adults 18 and over.
This test is part of Calqora's full Face Analysis tool: the same photo also gives you scores for every other category, so you get the complete picture, not just one number.
How We Calculate Your Score
No black box: your Harmony Score is a weighted composite of five measurable categories, each computed from a 478-point 3D face mesh detected on your device. The feature-by-feature breakdown of your eyes, eyebrows, nose, and lips shown with the report is descriptive detail. It does not feed into the score.
Facial Symmetry
weight 25%
A midline is drawn through your nose bridge and chin, then 9 landmark pairs (eye corners, brow ends, nostrils, mouth corners, jaw contour) are compared across it, both horizontal distance and vertical height.
Facial Thirds & Proportions
weight 25%
Your face is split into vertical thirds (hairline→brow, brow→nose base, nose base→chin) and scored on how evenly they divide (60%), plus your face length ÷ cheekbone width against the classical 1.618 golden-ratio reference (40%).
Eye Spacing & Canthal Tilt
weight 20%
The distance between your inner eye corners is compared to the width of one eye (the classical rule of fifths), and the angle from outer to inner eye corner measures canthal tilt.
Jawline Definition
weight 15%
Jaw width ÷ cheekbone width measures taper (60%), and the angle formed at your chin between the left and right jaw contours measures chin definition (40%).
Cheekbone Prominence
weight 15%
Cheekbone width ÷ jaw width (60%) and cheekbone width ÷ face length (40%) measure how strongly the midface projects relative to the lower face.
Related Tools
Face Shape Detector →
Identify your face shape (oval, round, square, heart, diamond, oblong, or triangle) with hairstyle and contouring guidance.
AI Glasses Finder →
Frame recommendations matched to your face shape, with visual previews of what suits you and what to avoid.
Height Comparison →
Compare heights visually side by side: the tool that started Calqora.
Frequently Asked Questions
How accurate is this attractiveness test?
The measurements are accurate to within a few pixels on a clear, straight-on photo. What they predict is a different question. Symmetry and averageness have real research support as attractiveness factors; classical formulas like the golden ratio have mixed support at best. So the geometry is measured precisely and reported honestly, but no proportion test can tell you how attractive other people find you.
What is a good score on an attractiveness test?
On this tool most clear front-facing photos land between about 70 and 90, and the scale deliberately floors at 58 rather than 0 because differences below that resolution would be false precision from a 2D photo. Above roughly 88 means your measured proportions sit close to the classical references across every category. Below that usually means one or two categories diverge, which is the normal condition of a real face.
Why won't this tool just rate my face out of 10?
Because a single rating implies a precision and an authority that no photo-based measurement has. A number with no method behind it is also the format most likely to make someone feel worse about a face that is completely unremarkable in the statistical sense. This tool reports what it measured, what the reference value is, and what you can actually do with that information in styling terms.
Does the test work the same for everyone?
The landmark detection and every ratio are computed identically regardless of sex, gender, or ethnicity, and everyone is scored against one neutral set of classical references. Those references originate in European art convention, which is a genuine limitation worth naming: they describe one tradition of proportion, not a universal standard of beauty. That is another reason the output is framed as styling guidance rather than a verdict.
Is my photo uploaded for this attractiveness test?
No. The AI model downloads to your browser and every step, detection, scoring, and the mirrored composites, runs on your device. Your photo is never transmitted, stored, or logged, and the shareable result card contains only your scores. Close the tab and the image is gone.